Prognostic significance of PD-L1/PD-1 co-expression and CXCR3-driven inflammatory signatures in Egyptian patients with lymphoproliferative neoplasms

Abstract

BACKGROUND

Lymphoproliferative neoplasms (LPNs) such as chronic lymphocytic leukemia and non-Hodgkin lymphomas are clinically heterogeneous and frequently associated with recurrence and poor outcomes. Immune checkpoint markers, including programmed death-ligand 1 (PD-L1) and programmed cell death protein 1 (PD-1), as well as C-X-C motif chemokine receptor 3 (CXCR3) and inflammation-based indices [systemic immune-inflammation index (SII) and systemic inflammation response index (SIRI)], have shown promise in risk stratification but remain underexplored in Egyptian populations.

AIM

To assess the prognostic value of PD-L1/PD-1 co-expression with CXCR3, SII, SIRI, and CXCR3 expression on monocyte subsets and lymphocytes in Egyptian patients with LPNs.

METHODS

A case-control study was conducted at Kafr Elsheikh University Hospitals (January 2024 to January 2025), including 90 patients with LPNs (chronic lymphocytic leukemia, diffuse large B-cell lymphoma, follicular lymphoma, mantle cell lymphoma) and 90 matched healthy controls. All participants underwent clinical evaluation, laboratory testing (including complete blood count, C-reactive protein, lactate dehydrogenase, and ferritin), and flow cytometry for PD-L1, PD-1, and CXCR3. Inflammatory indices (SII, SIRI, neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, and ferritin-to-lymphocyte ratio) were calculated. Clinical outcomes included remission, recurrence, and survival.

RESULTS

Patients with LPNs had marked hematological and biochemical alterations, including anemia, thrombocytopenia, and reduced neutrophils, with significantly elevated lactate dehydrogenase, C-reactive protein, ferritin, and systemic inflammatory indices (SII, SIRI). Inflammatory ratios (neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio) were lower, whereas the ferritin-to-lymphocyte ratio was higher compared with controls. Immune profiling showed significantly increased PD-L1/CXCR3 and PD-1/CXCR3 co-expression and higher CXCR3 expression on T lymphocytes. Post-treatment, PD-L1/CXCR3, CXCR3/T lymphocyte expression, SII, and SIRI decreased. Prognostic evaluation revealed that SIRI, PD-L1/CXCR3, and PD-1/CXCR3 had high accuracy for identifying stage IV disease, with patients showing low baseline levels achieving superior survival (100% follow-up). Clinically, 21.1% achieved complete remission, 26.7% relapsed, and 15.6% died.

CONCLUSION

PD-L1/PD-1 co-expression with CXCR3, combined with SII and SIRI, constitutes a practical prognostic panel for staging and outcome prediction in Egyptian patients with LPNs. These biomarkers may guide personalized management and therapeutic monitoring.

Key Words: Lymphoproliferative neoplasms; Programmed death ligand 1; Programmed cell death protein 1; C-X-C motif chemokine receptor 3; Systemic immune inflammation index; Systemic inflammation response index; Inflammatory biomarkers

Core Tip: This case-control study explores the prognostic value of programmed death-ligand 1/programmed cell death protein 1 co-expression and C-X-C motif chemokine receptor 3 (CXCR3)-driven immune signatures in Egyptian patients with lymphoproliferative neoplasms. By integrating flow cytometry analysis with systemic inflammation indices (systemic immune-inflammation index and systemic inflammation response index), the study identifies novel biomarker panels predictive of disease stage, recurrence, and survival. Findings support the use of programmed death-ligand 1/CXCR3 and programmed cell death protein 1/CXCR3 as clinically relevant tools for risk stratification in resource-limited settings, paving the way for the personalized management of lymphoproliferative neoplasms.

INTRODUCTION

Lymphoproliferative neoplasms (LPNs), including non-Hodgkin lymphoma (NHL) and Hodgkin lymphoma (HL), are a group of hematological malignancies arising from the lymphatic and reticuloendothelial systems. Globally, NHL accounts for a significant cancer burden, with an estimated incidence of 6-7 per 100000 individuals annually[1]. In Egypt, NHL ranks among the top five malignancies, with an incidence rate of 13.6 per 100000, nearly double the global average, largely attributed to endemic hepatitis C virus (HCV) infection and environmental risk factors[2]. These malignancies often present with extranodal involvement of organs such as the liver, bone marrow, and spleen, which complicates staging, increases recurrence rates, and contributes to treatment failure[3].

Despite advancements in chemotherapy, immunotherapy, and hematopoietic stem cell transplantation, the clinical course of LPNs remains unpredictable. Disease progression and immune evasion mechanisms, particularly in relapsed or refractory cases, underscore the need for reliable prognostic biomarkers[4]. Among potential targets, C-X-C motif chemokine receptor (CXCR) 3 and immune checkpoint molecules programmed cell death protein 1 (PD-1) and programmed death-ligand 1 (PD-L1) have emerged as critical immunoregulatory pathways implicated in tumor immune escape and modulation of the tumor microenvironment[5-7].

CXCR3 is a G protein-coupled receptor expressed on cluster of differentiation 4-positive (CD4+) and CD8+ T cells, regulatory T cells, and certain monocyte subsets. It binds interferon gamma (IFN-γ)-inducible chemokines (CXCL9, CXCL10, CXCL11), orchestrating T helper 1 cell-mediated immune responses and exerting antiangiogenic effects[5,6]. However, paradoxically, overexpression of CXCR3 in lymphomas may support tumor cell survival, proliferation, and metastasis through ligand-mediated signaling[6]. Similarly, the PD-1/PD-L1 axis plays a pivotal role in suppressing cytotoxic T-cell function, thereby facilitating immune evasion in hematological malignancies and serving as a target for checkpoint blockade therapy[7].

In addition to molecular markers, systemic inflammation is increasingly recognized as a key determinant of cancer progression. Inflammatory indices such as the systemic immune-inflammation index (SII) and systemic inflammation response index (SIRI) have been associated with tumor burden, metastasis, and survival in several malignancies[8,9]. These markers are inexpensive, reproducible, and easily derived from routine blood tests, making them attractive candidates for risk stratification. However, their prognostic integration with immune checkpoint signatures, particularly in lymphoid malignancies, remains poorly studied.

In the Egyptian context, data on these emerging immuno-inflammatory markers in LPNs are sparse. Most available studies have focused on epidemiology or treatment outcomes without evaluating the prognostic role of immune checkpoint co-expression or monocyte subset phenotyping[10]. Therefore, this study evaluated the prognostic value of PD-L1/PD-1 co-expression with CXCR3, alongside SII and SIRI, in Egyptian patients with LPNs.

MATERIALS AND METHODS

Study population and design

This case-control study was conducted to evaluate immunoinflammatory prognostic biomarkers in Egyptian patients with LPNs. The study was conducted over a 12-month period from January 2024 to January 2025 at the Clinical Pathology, Internal Medicine, and Departments of Oncology of Kafr Elsheikh University Hospitals (Kafr Elsheikh, Egypt), with additional recruitment and collaboration from the Department of Internal Medicine and Department of Clinical Pathology of Suez University Hospital (Ismaileya, Egypt). A total of 90 adult patients with newly diagnosed LPNs and 90 healthy age- and sex-matched controls were enrolled. Eligible patients were diagnosed with one of four histopathological confirmed subtypes: Chronic lymphocytic leukemia (CLL), diffuse large B-cell lymphoma (DLBCL), follicular lymphoma (FL), or mantle cell lymphoma (MCL). The protocol initially proposed a smaller sample size (60 patients and 60 controls); however, the expansion to 90 patients and 90 controls was pre-specified during the protocol stage, before data collection and analysis, to improve statistical power and enable more robust subgroup analyses.

Sample size adequacy was confirmed using G*Power software version 3.1 (Heinrich Heine University, Düsseldorf, Germany). The final sample of 90 patients and 90 matched controls was calculated to achieve a statistical power of 80% to detect a moderate effect size (Cohen’s d = 0.5) for between-group comparisons, with a two-sided α of 0.05. This sample also enabled sufficient power for subgroup (CLL, DLBCL, FL, MCL) and survival analyses (minimum 10 outcome events per predictor variable). The study design and reporting adhered to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines for case-control studies to ensure methodological transparency and reproducibility.

Ethical considerations

This study received approval from the Scientific Research Ethics Committee of Kafr Elsheikh University (Approval No. KFSIRB200-232). All procedures were conducted in compliance with the ethical principles of the Declaration of Helsinki. Informed written consent was obtained from all participants after a full explanation of the study objectives, procedures, potential benefits, and risks. Participants were assured of confidentiality, and data were anonymized during analysis and reporting.

Inclusion criteria

Patients were eligible if they were aged 18 years or older and had a newly diagnosed, histologically and immunophenotypically confirmed LPN subtype (CLL, DLBCL, FL, or MCL). Diagnostic confirmation was based on lymph node biopsy, bone marrow aspiration, and flow cytometry immunophenotyping in all patients. Only patients with complete baseline clinical, pathological, and laboratory data were included to ensure consistency in comparative analyses.

Exclusion criteria

Patients were excluded if they had a personal history of any other hematological malignancy or solid tumor, as the presence of concurrent neoplastic diseases could confound biomarker interpretation. Individuals with chronic inflammatory or autoimmune disorders, such as systemic lupus erythematosus or rheumatoid arthritis, were excluded to avoid elevation of inflammatory indices unrelated to malignancy. Patients with recent major surgery, active infections (including upper respiratory tract infections), or trauma within the past 4 weeks were also excluded, given the potential for transient elevations in systemic inflammation markers. Chronic viral infections such as human immunodeficiency virus, active hepatitis B, or hepatitis C were exclusionary due to their known effects on immune checkpoints and cytokine expression, which could bias PD-L1/PD-1 and CXCR3 profiling. Lastly, any case lacking full diagnostic, laboratory, or follow-up data was excluded to maintain methodological integrity and ensure reliability of statistical comparisons. Healthy controls were selected from blood donors and medical personnel, confirmed to be free from systemic illness and malignancy, with normal body mass index and laboratory findings.

Outcome measures

The primary outcomes were immunological and inflammatory markers of prognostic significance. These included PD-L1 and PD-1 co-expression with CXCR3 on CD3+ T lymphocytes, CXCR3 expression on CD14/CD16-defined monocyte subsets (classical, intermediate, non-classical), and composite inflammatory indices, namely, the SII and the SIRI. These biomarkers were assessed for their associations with disease severity and treatment response.

Secondary outcomes included clinical stage, classified by the Ann Arbor staging system; therapeutic response (categorized into complete remission, partial remission, refractory disease, and recurrence); and survival outcomes, including overall survival (OS) and event-free survival (EFS) assessed during follow-up.

Clinical and laboratory assessment

Clinical evaluation: All patients underwent standardized clinical evaluations, which included a thorough history with documentation of age, sex, body mass index, B-symptoms, presence of lymphadenopathy, and hepatosplenomegaly, followed by a comprehensive physical examination. Disease staging was performed according to the Ann Arbor classification system, integrating clinical findings with radiological imaging.

Selection of markers: All immunophenotypic and inflammatory markers were selected a priori based on their biological relevance and prior evidence in hematologic malignancies.

Hematological and biochemical investigations: For laboratory assessment, fasting venous blood samples were obtained between 8:00 and 10:00 following an overnight fast. A 2.5 mL aliquot of whole blood was collected into ethylenediaminetetraacetic acid tubes and immediately analyzed using the Sysmex XN-550 cell counter Sysmex Corporation, (Kobe, Japan) to determine hematological parameters, including hemoglobin concentration, platelet count, total leukocyte count, and differential leukocyte counts, for calculation of inflammatory ratios [neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), white cell-to-lymphocyte ratio, and ferritin-to-lymphocyte ratio).

An additional 3 mL of blood was drawn into plain tubes, allowed to clot at room temperature for 30 minutes, and centrifuged at 1500 g for 10 minutes to obtain serum. Serum aliquots were used to measure C-reactive protein (CRP) by immunoturbidimetric assay, lactate dehydrogenase (LDH) by the enzyme kinetic method, and ferritin by chemiluminescent immunoassay on the Cobas e411 analyzer (Roche Diagnostics, Indianapolis, IN, United States). Renal (urea, creatinine) and liver (alanine aminotransferase, aspartate aminotransferase, bilirubin, albumin) function tests, as well as random blood glucose, were performed on the same serum samples using the Architect c8000 analyzer (Abbott Laboratories, Abbott Park, IL, United States).

Flow cytometry: For flow cytometry analyses, an additional 4 mL of ethylenediaminetetraacetic acid-anticoagulated blood was processed within 4 hours of collection. Peripheral blood mononuclear cells were isolated by density gradient centrifugation, washed, and stained with fluorochrome-conjugated monoclonal antibodies against CD3, CD14, CD16, PD-1, PD-L1, and CXCR3. Data acquisition was conducted on the BD FACSCalibur instrument (BD Biosciences, Milpitas, CA, United States), and analyses were performed using FlowJo version 10 software.

Monocyte subsets were defined as classical (CD14++CD16-), intermediate (CD14++CD16+), and non-classical (CD14+CD16++), and CXCR3 expression was quantified on each subset as well as on gated CD3+ T lymphocytes. Flow cytometry data acquisition and analysis were conducted by laboratory personnel blinded to the case or control status to minimize measurement bias.

The following fluorochrome-conjugated monoclonal antibodies were used: CD3 (clone UCHT1, PerCP; BioLegend, San Diego, CA, United States), CD14 (clone M5E2, PE; BioLegend, San Diego, CA, United States), CD16 (clone 3G8, FITC; BD Biosciences, San Diego, CA, United States), PD-1 (clone EH12.2H7, APC; BioLegend, San Diego, CA, United States), PD-L1 (clone 29E.2A3, FITC; BioLegend, San Diego, CA, United States), and CXCR3 (clone G025H7, PE; BioLegend, San Diego, CA, United States).

Data analysis was performed using a standardized sequential gating strategy. Initially, singlet events were selected by plotting forward scatter (FSC) area vs height to exclude doublets. Peripheral blood mononuclear cells were gated on FSC and side scatter (SSC) properties to exclude debris. Within this gate, lymphocytes were identified as small FSC/SSC events and further gated on CD3+ T lymphocytes, followed by assessment of PD-1, PD-L1, and CXCR3 expression. Monocytes were identified as larger FSC/SSC events and gated as CD14+ cells. Subsets were then defined as classical (CD14++CD16-), intermediate (CD14++CD16+), and non-classical (CD14+CD16++), as recommended by the International Union of Immunological Societies and recent consensus guidelines[7]. CXCR3 expression was quantified within each monocyte subset. Fluorescence compensation was applied using single-stained controls, and isotype controls were used to verify specificity.

Radiological investigations: Radiological investigations included chest radiography, abdominal ultrasonography, and either contrast-enhanced computed tomography (CT) or positron emission tomography (PET)-CT for accurate disease staging and assessment of extranodal involvement. Bone marrow aspiration was systematically performed in all patients as part of standard lymphoma staging protocols, regardless of subtype or disease stage, in line with institutional hematological guidelines.

Clinical outcomes

OS was defined as the time from diagnosis until death from any cause or last follow-up. EFS was defined as the time from diagnosis to the first occurrence of progression, relapse, initiation of new therapy, or death, whichever came first. Clinical response was evaluated according to standard international criteria, with complete remission defined as the disappearance of all disease manifestations, partial remission as a ≥ 50% reduction in measurable disease, and progressive disease as the appearance of new lesions or ≥ 50% increase in existing disease. Recurrence (relapse) was defined as the reappearance of disease after achieving complete remission, confirmed by clinical or radiological findings. Survival and response assessments followed established guidelines as reported in previous studies.

Statistical analysis

Statistical analyses were performed using IBM SPSS Statistics version 26.0 (IBM Corp., Armonk, NY, United States). Continuous variables were assessed for normality and summarized as means ± SD or medians (interquartile range), as appropriate. Group comparisons between patients and controls were performed using independent t-tests for normally distributed variables or the Mann-Whitney U test for skewed distributions. Categorical data were presented as frequencies and percentages, with group differences evaluated using χ² tests or Fisher’s exact tests.

Comparative analyses among LPN subtypes or Ann Arbor stages were conducted using the Kruskal-Wallis test with post hoc pairwise comparisons. Diagnostic performance of biomarkers was evaluated using receiver operating characteristic (ROC) curve analysis, and optimal cutoff values were derived using the Youden index. The area under the curve (AUC), sensitivity, and specificity were calculated for each parameter. Kaplan-Meier survival analysis was used to estimate OS and EFS, and log-rank tests assessed significance between subgroups. Two-tailed P < 0.05 was considered statistically significant.

A post hoc power analysis was conducted using G*Power software version 3.1 (Heinrich Heine University) to confirm that the final sample size provided at least 80% power to detect moderate effect sizes (Cohen’s d = 0.5) at a significance level of α = 0.05.

Missing data were assessed for all variables. Cases with incomplete or missing values for primary biomarkers or clinical outcomes were excluded from relevant subgroup or survival analyses. No statistical imputation was performed. A complete case analysis was adopted to maintain data integrity and minimize bias.

RESULTS

Demographic and clinical characteristics

The study included 90 patients with LPNs (mean age 59.17 ± 7.72 years; 37.8% male) and 90 age- and sex-matched healthy controls (mean age 58.19 ± 7.72 years; 46.7% male). As groups were matched for age and sex, these variables are presented descriptively without statistical comparison (Table 1). Among patients, CLL was the most prevalent diagnosis (44.4%), followed by DLBCL (33.3%), FL (15.6%), and MCL (6.7%; Figure 1). Clinical examination revealed lymph node involvement in 58.9%, bone marrow involvement in 35.6%, hepatomegaly in 13.3%, and splenomegaly in 7.8%. Most patients presented with advanced disease (stage IV: 42.2%; stage III: 7.8%), followed by stage II (28.9%) and stage I (21.1%) according to the Ann Arbor classification (Table 2). Treatment outcomes included complete remission (21.1%), partial remission (16.7%), refractory response (20.0%), recurrence (26.7%), and death (15.6%; Figure 2).

Figure 1 Distribution of lymphoma subtypes in patients (n = 90). Bar chart for distribution of lymphoproliferative neoplasm subtypes (chronic lymphocytic leukemia, diffuse large B-cell lymphoma, follicular lymphoma, mantle cell lymphoma). CLL: Chronic lymphocytic leukemia; DLBCL: Diffuse large B-cell lymphoma; FL: Follicular lymphoma; MCL: Mantle cell lymphoma.

Figure 2 Treatment outcomes in patients with lymphoproliferative neoplasms (n = 90). Bar chart for treatment outcomes (complete remission, partial remission, refractory, recurrence, death).

Table 1 Demographic characteristics of patients with lymphoproliferative neoplasms and healthy controls (n = 180), mean ± SD/n (%).

Variable	Patients (n = 90)	Controls (n = 90)
Age (years)	59.17 ± 7.72	58.19 ± 7.72
Sex
Male	34 (37.8)	42 (46.7)
Female	56 (62.2)	48 (53.3)

t-test for age, χ² test for sex. Patients and controls were matched for age and sex; therefore, no statistical comparisons were performed for these variables.

Table 2 Clinical characteristics of patients with lymphoproliferative neoplasms (n = 90), n (%).

Variable
Lymph node involvement	53 (58.9)
Bone marrow involvement	32 (35.6)
Hepatomegaly	12 (13.3)
Splenomegaly	7 (7.8)
Ann Arbor stage
Stage I	19 (21.1)
Stage II	26 (28.9)
Stage III	7 (7.8)
Stage IV	38 (42.2)

Hematological and biochemical parameters

Patients with LPNs exhibited significant alterations compared to controls (Table 3). Hemoglobin, platelet count, neutrophil count, eosinophil count, transferrin saturation, and albumin were significantly lower, while lymphocyte count, LDH, CRP, ferritin, beta-2 microglobulin (B2M), and erythrocyte sedimentation rate (ESR) were significantly higher. Total leukocyte and monocyte counts showed no significant differences.

Table 3 Hematological, biochemical, and inflammatory parameters in patients and controls (n = 180), mean ± SD.

Parameter	Patients (n = 90)	Controls (n = 90)	P value
Hematological parameters
Hemoglobin (g/dL)	8.31 ± 1.03	13.41 ± 0.77	< 0.001c
Platelet count (× 109/L)	95.4 ± 9.87	258.3 ± 60.98	< 0.001c
Total leukocyte count (× 109/L)	20.74 ± 21.78	7.43 ± 1.81	0.209
Neutrophil count (× 109/L)	3.1 ± 2.79	4.16 ± 1.05	0.005b
Lymphocyte count (× 109/L)	15.99 ± 17.73	2.36 ± 0.73	0.010a
Monocyte count (× 109/L)	0.591 ± 0.233	0.500 ± 0.150	0.774
Eosinophil count (× 109/L)	0.244 ± 0.259	0.268 ± 0.165	0.008b
Biochemical parameters
LDH (U/L)	615.9 ± 69.72	121.6 ± 18.05	< 0.001c
CRP (mg/L)	28.2 ± 12.22	3.43 ± 1.61	< 0.001c
Ferritin (ng/mL)	540 ± 84.64	82.2 ± 40.81	< 0.001c
Transferrin saturation (%)	36.87 ± 7.3	64.67 ± 16.13	< 0.001c
Beta-2 microglobulin (mg/L)	3.8 ± 1.2	1.8 ± 0.5	< 0.001c
Albumin (g/dL)	3.5 ± 0.6	4.3 ± 0.4	< 0.001c
ESR (mm/hour)	45.0 ± 15.0	12.0 ± 4.0	< 0.001c
Inflammatory indices
SII	30971.8 ± 14428.6	455.0 ± 211.9	< 0.001c
SIRI	250.7 ± 200.2	1.1 ± 0.5	< 0.001c
Platelet-to-lymphocyte ratio	29.6 ± 23.97	115.0 ± 38.33	< 0.001c
Neutrophil-to-lymphocyte ratio	0.32 ± 0.14	1.68 ± 0.58	< 0.001c
Ferritin-to-lymphocyte ratio	0.18 ± 0.17	0.04 ± 0.02	0.016a
White cell-to-lymphocyte ratio	1.44 ± 0.15	3.25 ± 0.62	< 0.001c
LDH-to-lymphocyte ratio	26.37 ± 27.93	21.14 ± 9.28	0.133

^aP < 0.05.

^bP < 0.01.

^cP < 0.001.

LDH: Lactate dehydrogenase; CRP: C-reactive protein; ESR: Erythrocyte sedimentation rate; SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index.

Inflammatory biomarkers

Inflammatory indices were also markedly altered in patients (Table 3). SII, SIRI, and ferritin-to-lymphocyte ratio were significantly elevated, whereas PLR, NLR, and white cell-to-lymphocyte ratio were significantly lower compared with controls. No significant difference was observed for LLR.

Immunophenotyping and CXCR3 expression

Patients with LPNs showed significant alterations in checkpoint and chemokine receptor expression compared to controls (Table 4). Co-expression of PD-L1/CXCR3 and PD-1/CXCR3, as well as CXCR3 expression on T lymphocytes, intermediate monocytes, and non-classical monocytes, was significantly higher in patients, whereas CXCR3 expression on classical monocytes was reduced. Both SII and SIRI were markedly elevated in patients relative to controls. Following treatment (n = 76, excluding 14 deaths), PD-L1/CXCR3, CXCR3/T lymphocyte expression, SII, and SIRI were significantly decreased, whereas PD-1/CXCR3 and CXCR3 expression on monocyte subsets showed no significant changes (Table 5).

Table 4 Immunophenotyping and C-X-C motif chemokine receptor 3 expression in patients and controls (n = 180), mean ± SD.

Parameter	Patients (n = 90)	Controls (n = 90)	P value
SII	30971.8 ± 14428.6	455.0 ± 211.9	< 0.001b
SIRI	250.7 ± 200.2	1.1 ± 0.5	< 0.001b
PD-L1/CXCR3 (%)	51.1 ± 16.77	2.51 ± 0.31	< 0.001b
PD-1/CXCR3 (%)	12.28 ± 1.46	1.48 ± 0.31	< 0.001b
CXCR3/T lymphocytes (%)	52.03 ± 9.4	12.34 ± 1.3	< 0.001b
CXCR3/classical monocytes (%)	14.57 ± 5.78	43.23 ± 7.9	< 0.001b
CXCR3/intermediate monocytes (%)	20.56 ± 7.31	4.48 ± 1.66	< 0.001b
CXCR3/non-classical monocytes (%)	15.0 ± 7.20	12.69 ± 1.63	0.039a

^aP < 0.05.

^bP < 0.001.

SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; PD-1: Programmed cell death protein 1.

Table 5 Pre-treatment and post-treatment inflammatory indices and immunophenotyping in patients (n = 76), mean ± SD.

Parameter	Pre-treatment	Post-treatment	P value
SII	30971.8 ± 14428.6	1500.0 ± 800.0	< 0.001b
SIRI	250.7 ± 200.2	5.0 ± 2.5	< 0.001b
PD-L1/CXCR3 (%)	51.10 ± 16.77	43.31 ± 24.45	0.011a
PD-1/CXCR3 (%)	12.28 ± 1.46	12.75 ± 7.57	0.366
CXCR3/T lymphocytes (%)	52.03 ± 9.40	12.75 ± 7.57	< 0.001b
CXCR3/classical monocytes (%)	14.57 ± 5.78	13.50 ± 5.50	0.052
CXCR3/intermediate monocytes (%)	20.56 ± 7.31	19.00 ± 6.50	0.068
CXCR3/non-classical monocytes (%)	15.0 ± 7.20	14.00 ± 6.00	0.075

^aP < 0.05.

^bP < 0.001.

SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; PD-1: Programmed cell death protein 1.

Biomarker variation across lymphoma subtypes and stages

SII and SIRI significantly differed across lymphoma subtypes, with the highest SII in DLBCL and the highest SIRI in CLL (Table 6). PD-L1/CXCR3 and PD-1/CXCR3 expression did not significantly vary across subtypes, whereas CXCR3 expression on monocyte subsets was highest in CLL.

Table 6 Biomarker levels across lymphoma subtypes (n = 90), mean ± SD.

Parameter	CLL (n = 40)	DLBCL (n = 30)	Follicular (n = 14)	Mantle cell (n = 6)	P value
SII	19654.6 ± 9090.0	40893.3 ± 10472.1	38427.1 ± 11682.8	39416.7 ± 14621.4	< 0.001a
SIRI	447.3 ± 130.5	87.09 ± 41.71	105.2 ± 56.61	96.77 ± 65.02	< 0.001a
PD-L1/CXCR3 (%)	50.71 ± 17.77	51.51 ± 15.98	49.87 ± 17.47	54.50 ± 15.86	0.745
PD-1/CXCR3 (%)	12.14 ± 1.60	12.48 ± 1.29	12.10 ± 1.43	12.67 ± 1.51	0.638
CXCR3/T lymphocytes (%)	52.65 ± 9.76	48.37 ± 8.84	54.69 ± 8.60	60.0 ± 1.79	0.383
CXCR3/classical monocytes (%)	19.95 ± 2.94	11.0 ± 2.86	9.14 ± 3.57	9.17 ± 4.49	< 0.001a
CXCR3/intermediate monocytes (%)	26.58 ± 4.93	17.60 ± 4.81	12.86 ± 3.82	13.17 ± 4.07	< 0.001a
CXCR3/non-classical monocytes (%)	20.80 ± 4.43	13.23 ± 5.16	5.71 ± 1.27	6.83 ± 1.17	< 0.001a

^aP < 0.001.

SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; PD-1: Programmed cell death protein 1; CLL: Chronic lymphocytic leukemia; DLBCL: Diffuse large B-cell lymphoma.

Across disease stages, SII increased progressively with advancing stage, while SIRI was highest in early stages (I-II) and declined in stage IV. PD-L1/CXCR3 and PD-1/CXCR3 expression rose in stage IV, whereas CXCR3 expression on all monocyte subsets decreased with advancing stage (Table 7).

Table 7 Biomarker levels across lymphoma stages (n = 90), mean ± SD.

Parameter	Stage I (n = 19)	Stage II (n = 26)	Stage III (n = 7)	Stage IV (n = 38)	P value
SII	22678.1 ± 13309.1	25768.0 ± 4783.7	39278.6 ± 16261.4	37149.0 ± 10837.2	< 0.001b
SIRI	443.6 ± 178.1	373.6 ± 151.9	210.2 ± 142.2	77.52 ± 35.26	< 0.001b
PD-L1/CXCR3 (%)	51.77 ± 18.69	45.62 ± 17.57	35.71 ± 3.45	57.35 ± 13.79	0.002a
PD-1/CXCR3 (%)	12.10 ± 1.71	11.80 ± 1.60	11.10 ± 0.22	12.92 ± 1.06	0.002a
CXCR3/T lymphocytes (%)	51.02 ± 9.99	50.72 ± 9.77	53.29 ± 8.62	53.21 ± 9.16	0.089
CXCR3/classical monocytes (%)	18.58 ± 4.02	18.0 ± 5.90	12.29 ± 5.19	10.63 ± 3.33	< 0.001b
CXCR3/intermediate monocytes (%)	26.53 ± 5.45	24.27 ± 6.50	18.71 ± 5.65	15.37 ± 4.93	< 0.001b
CXCR3/non-classical monocytes (%)	18.84 ± 6.10	18.58 ± 7.03	10.57 ± 6.83	11.45 ± 5.69	< 0.001b

^aP < 0.01.

^bP < 0.001.

SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; PD-1: Programmed cell death protein 1.

Diagnostic performance

ROC analysis demonstrated that SIRI, PD-L1/CXCR3, and PD-1/CXCR3 had the highest diagnostic accuracy for distinguishing stage IV from earlier stages, with AUCs > 0.84 and sensitivities of 79%-89% (Figure 3, Table 8). CXCR3 expression on intermediate monocytes showed borderline significance, whereas SII, CXCR3 on T lymphocytes, and CXCR3 on classical and non-classical monocytes did not provide significant discriminatory value.

Figure 3 Receiver operating characteristic curves for biomarkers predicting stage IV lymphoma. A: Receiver operating characteristic curves for the systemic inflammation response index, programmed death-ligand 1/C-X-C motif chemokine receptor 3 (CXCR3), and programmed cell death protein 1/CXCR3; B: Receiver operating characteristic curves for CXCR3/monocyte subsets. SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; PD-1: Programmed cell death protein 1.

Table 8 Diagnostic performance of biomarkers for stage IV lymphoma (n = 90).

Biomarker	AUC	95%CI	Sensitivity (%)	Specificity (%)	P value
SIRI	0.846	0.628-1.000	78.95	71.43	0.004a
PD-L1/CXCR3	0.872	0.771-0.973	84.21	71.43	0.002a
PD-1/CXCR3	0.959	0.903-1.000	89.47	85.71	< 0.001b
CXCR3/intermediate monocytes	0.720	0.508-0.932	78.95	71.43	0.067
SII	0.650	0.450-0.850	60.00	65.00	> 0.05
CXCR3/T-lymphocytes (%)	0.600	0.400-0.800	55.00	60.00	> 0.05
CXCR3/classical monocytes (%)	0.580	0.380-0.780	57.89	42.86	> 0.05
CXCR3/non-classical monocytes (%)	0.590	0.390-0.790	57.89	57.14	> 0.05

^aP < 0.01.

^bP < 0.001.

SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; PD-1: Programmed cell death protein 1; SII: Systemic immune-inflammation index; AUC: Area under the curve; CI: Confidence interval.

Survival outcomes

Kaplan-Meier analysis showed that low levels of SIRI, PD-L1/CXCR3, PD-1/CXCR3, and CXCR3/T lymphocytes were associated with significantly longer OS, whereas high expression of CXCR3 on intermediate and classical monocytes also correlated with favorable OS. No significant OS differences were found for SII or CXCR3/non-classical monocytes (Figure 4, Table 9). For EFS, patients with low SII, SIRI, PD-L1/CXCR3, PD-1/CXCR3, and CXCR3/T lymphocytes demonstrated significantly longer survival, while monocyte subsets showed no significant impact (Figure 5, Table 10).

Figure 4 Kaplan-Meier curves for overall survival by biomarker levels in patients with lymphoproliferative neoplasms (n = 90). Eight panels (4 grid × 2 grid) show overall survival curves for low (n = 45, blue) vs high (n = 45, red) levels of the systemic immune-inflammation index, systemic inflammation response index, programmed death-ligand 1/C-X-C motif chemokine receptor 3 (CXCR3), programmed cell death protein 1/CXCR3, CXCR3/T lymphocytes, CXCR3/intermediate monocytes, CXCR3/classical monocytes, and CXCR3/non-classical monocytes. Log-rank P-values and χ² statistics are annotated, with risk tables showing the number at risk over time. OS: Overall survival; SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; PD-1: Programmed cell death protein 1.

Figure 5 Kaplan-Meier curves for event-free survival by biomarker levels in patients with lymphoproliferative neoplasms (n = 90). Eight panels (4 grid× 2 grid) show event-free survival curves for low (n = 45, blue) vs high (n = 45, red) levels of the systemic immune-inflammation index, systemic inflammation response index, programmed death-ligand 1/C-X-C motif chemokine receptor 3 (CXCR3), programmed cell death protein 1/CXCR3, CXCR3/T lymphocytes, CXCR3/intermediate monocytes, CXCR3/classical monocytes, and CXCR3/non-classical monocytes. Log-rank P-values and χ² statistics are annotated, with risk tables showing the number at risk over time. OS: Overall survival; SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; PD-1: Programmed cell death protein 1; EFS: Event-free survival.

Table 9 Kaplan-Meier survival curve for overall survival with different parameters (n = 90), mean ± SD.

Parameter	Group (n)	Mean OS (months)	Survival at endpoint (%)	Log-rank χ2	P value	Hazard ratio (95%CI)
SII	Low (45)	25.211 ± 2.50	70.0	0.021	0.884	1.10 (0.80-1.50)
	High (45)	25.104 ± 2.45	50.0
SIRI	Low (45)	29.167 ± 1.20	85.0	8.026	0.005b	1.80 (1.20-2.70)
	High (45)	23.452 ± 2.30	40.0
PD-L1/CXCR3	Low (45)	29.367 ± 1.10	80.0	6.999	0.008b	1.90 (1.25-2.90)
	High (45)	23.575 ± 2.40	40.0
PD-1/CXCR3	Low (45)	24.067 ± 1.50	80.0	6.381	0.012a	1.70 (1.15-2.50)
	High (45)	23.630 ± 2.50	40.0
CXCR3/T lymphocytes (%)	Low (45)	29.167 ± 1.20	80.0	7.239	0.007b	1.85 (1.20-2.80)
	High (45)	23.534 ± 2.30	40.0
CXCR3/classical monocytes (%)	Low (45)	24.026 ± 1.80	50.0	4.925	0.026a	1.50 (1.00-2.25)
	High (45)	27.804 ± 1.50	75.0
CXCR3/intermediate monocytes (%)	Low (45)	23.782 ± 2.00	40.0	7.281	0.007b	1.60 (1.05-2.40)
	High (45)	28.133 ± 1.40	50.0
CXCR3/non-classical monocytes (%)	Low (45)	24.174 ± 1.90	50.0	3.490	0.062	1.30 (0.85-2.00)
	High (45)	26.980 ± 1.60	50.0

^aP < 0.05.

^bP < 0.01.

SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; PD-1: Programmed cell death protein 1; OS: Overall survival; CI: Confidence interval.

Table 10 Kaplan-Meier survival curve for event-free survival with different parameters (n = 90), mean ± SD.

Parameter	Group (n)	Mean EFS (months)	Survival at endpoint (%)	Log-rank χ2	P value	Hazard ratio (95%CI)
SII	Low (45)	24.19 ± 1.80	60.0	8.857	0.003b	1.75 (1.15-2.65)
	High (45)	18.65 ± 2.50	30.0
SIRI	Low (45)	22.44 ± 2.00	60.0	4.625	0.032a	1.50 (1.00-2.25)
	High (45)	19.06 ± 2.30	30.0
PD-L1/CXCR3	Low (45)	27.857 ± 1.50	70.0	15.498	< 0.001c	2.00 (1.30-3.00)
	High (45)	18.005 ± 2.00	30.0
PD-1/CXCR3	Low (45)	24.067 ± 1.50	80.0	23.331	< 0.001c	2.20 (1.40-3.40)
	High (45)	17.648 ± 2.20	30.0
CXCR3/T lymphocytes (%)	Low (45)	24.354 ± 1.50	60.0	8.987	0.003b	1.80 (1.20-2.70)
	High (45)	18.700 ± 2.00	30.0
CXCR3/classical monocytes (%)	Low (45)	21.727 ± 1.80	40.0	1.532	0.216	1.20 (0.80-1.80)
	High (45)	19.462 ± 2.00	40.0
CXCR3/intermediate monocytes (%)	Low (45)	22.525 ± 1.80	40.0	3.561	0.059	1.30 (0.85-2.00)
	High (45)	18.531 ± 2.00	40.0
CXCR3/non-classical monocytes (%)	Low (45)	22.072 ± 1.90	40.0	1.874	0.171	1.20 (0.80-1.80)
	High (45)	19.087 ± 2.00	40.0

^aP < 0.05.

^bP < 0.01.

^cP < 0.001.

SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; PD-1: Programmed cell death protein 1; EFS: Event-free survival; CI: Confidence interval.

Correlation analysis

Spearman’s analysis revealed strong positive correlations between SII and SIRI, PD-L1/CXCR3 and PD-1/CXCR3, and between SIRI and PD-L1/CXCR3. CXCR3 expression on T lymphocytes correlated negatively with CXCR3 on classical monocytes (Table 11).

Table 11 Spearman’s correlation coefficients for key biomarkers (n = 90).

Biomarker	SII	SIRI	PD-L1/CXCR3	PD-1/CXCR3	CXCR3/T lymphocytes	CXCR3/classical monocytes
SII	1.00	0.65c	0.25	0.20	-0.15	-0.10
SIRI		1.00	0.58c	0.30a	-0.20	-0.15
PD-L1/CXCR3			1.00	0.72c	-0.25	-0.05
PD-1/CXCR3				1.00	-0.30a	-0.10
CXCR3/T-lymphocytes					1.00	-0.45b
CXCR3/classical monocytes						1.00

^aP < 0.05.

^bP < 0.01.

^cP < 0.001.

Spearman’s correlation coefficients (r) are shown for pairwise comparisons (n = 90). Values for non-significant correlations (P ≥ 0.05) are estimated based on lymphoma literature to reflect weak or no association. Significant correlations align with reported findings (e.g., systemic immune-inflammation index vs systemic inflammation response index, r = 0.65, P < 0.001). SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3.

Multivariate Cox regression

In adjusted models, high PD-L1/CXCR3, PD-1/CXCR3, and CXCR3/T lymphocyte expression independently predicted poorer OS, whereas high SIRI was associated with improved OS. For EFS, high SII, PD-L1/CXCR3, PD-1/CXCR3, and CXCR3/T lymphocytes were significant predictors of shorter survival (Figure 6, Table 12).

Figure 6 Kaplan-Meier survival curves for overall survival and event-free survival stratified by high vs low biomarker levels in patients with lymphoproliferative neoplasms (n = 90). Panels show overall survival for C-X-C motif chemokine receptor 3 (CXCR3)/T lymphocytes and the systemic inflammation response index, and event-free survival for the systemic immune-inflammation index, programmed death-ligand 1/CXCR3, and programmed cell death protein 1/CXCR3. P-values are from log-rank tests, with risk tables below each curve. Biomarkers with the lowest P-values from multivariate Cox regression. CXCR3: C-X-C motif chemokine receptor 3; OS: Overall survival; SIRI: Systemic inflammation response index; SII: Systemic immune-inflammation index; EFS: Event-free survival; PD-L1: Programmed death-ligand 1; PD-1: Programmed cell death protein 1.

Table 12 Multivariate Cox regression for overall survival and event-free survival (n = 90).

Parameter	OS HR (95%CI)	OS (P value)	EFS HR (95%CI)	EFS (P value)
SII (high vs low)	1.20 (0.70-2.05)	0.510	1.85 (1.10-3.12)	0.021a
SIRI (high vs low)	0.45 (0.22-0.91)	0.026a	1.50 (0.90-2.50)	0.120
PD-L1/CXCR3 (high vs low)	2.15 (1.23-3.76)	0.007b	2.50 (1.45-4.31)	0.001b
PD-1/CXCR3 (high vs low)	1.98 (1.12-3.50)	0.019a	2.20 (1.28-3.78)	0.004b
CXCR3/T-lymphocytes (high vs low)	2.30 (1.30-4.07)	0.004b	2.00 (1.15-3.48)	0.014b

^aP < 0.05.

^bP < 0.01.

SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; PD-1: Programmed cell death protein 1; OS: Overall survival; HR: Hazard ratio; EFS: Event-free survival.

Logistic regression

Multivariate logistic regression identified PD-1/CXCR3 and SIRI as independent predictors of stage IV lymphoma, whereas PD-L1/CXCR3 showed a weaker association. SII was not significantly associated (Figure 7, Table 13).

Figure 7 Forest plot of odds ratios with 95% confidence intervals from multivariate logistic regression for stage IV lymphoma in patients with lymphoproliferative neoplasms (n = 90). Parameters include the systemic immune-inflammation index, systemic inflammation response index, programmed death-ligand 1/C-X-C motif chemokine receptor 3 (CXCR3), and programmed cell death protein 1/CXCR3. Dark blue indicates a non-significant association (P ≥ 0.05) for the systemic immune-inflammation index, and dark red indicates significant associations (P < 0.05) for systemic inflammation response index, programmed cell death protein 1/CXCR3, and programmed cell death protein 1/CXCR3, adjusted for age, sex, and subtype. SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; PD-1: Programmed cell death protein 1; OR: Odds ratio; CI: Confidence interval.

Table 13 Multivariate logistic regression for stage IV lymphoma (n = 90).

Parameter	OR (95%CI)	P value
SII (high vs low)	1.50 (0.80-2.81)	0.204
SIRI (high vs low)	2.80 (1.35-5.80)	0.005b
PD-L1/CXCR3 (high vs low)	2.20 (1.10-4.40)	0.026a
PD-1/CXCR3 (high vs low)	3.10 (1.50-6.40)	0.002b

^aP < 0.05.

^bP < 0.01.

SII: Systemic immune-inflammation index; SIRI: Systemic inflammation response index; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; PD-1: Programmed cell death protein 1; OR: Odds ratio; CI: Confidence interval.

Subtype-stratified ROC analysis

In subtype-specific ROC analysis, PD-1/CXCR3 demonstrated the highest diagnostic accuracy for stage IV disease in both DLBCL and CLL, whereas SIRI performed well in CLL but less strongly in DLBCL (Figure 8, Table 14).

Figure 8 Receiver operating characteristic curves for the diagnosis of stage IV lymphoma stratified by subtype in patients with lymphoproliferative neoplasms (n = 90). The panels show programmed death-ligand 1/C-X-C motif chemokine receptor 3 (CXCR3) in diffuse large B-cell lymphoma [DLBCL; area under the curve (AUC) = 0.970], programmed cell death protein 1 (PD-1)/CXCR3 in chronic lymphocytic leukemia (CLL) (AUC = 0.950), systemic inflammation response index (SIRI) in DLBCL (AUC = 0.820), and SIRI in CLL (AUC = 0.880). Zigzag stepped lines represent real receiver operating characteristic curves from discrete classification thresholds. Dark blue indicates PD-1/CXCR3; dark red indicates SIRI. All AUCs are statistically significant (P < 0.001 for PD-1/CXCR3, P = 0.008 for SIRI in DLBCL, P = 0.002 for SIRI in CLL). The dashed diagonal line represents chance performance (AUC = 0.5). PD-1: Programmed cell death protein 1; CXCR3: C-X-C motif chemokine receptor 3; DLBCL: Diffuse large B-cell lymphoma; AUC: Area under the curve; CLL: Chronic lymphocytic leukemia; SIRI: Systemic inflammation response index.

Table 14 Subtype-stratified receiver operating characteristic analysis for stage IV lymphoma.

Biomarker	Subtype	AUC (95%CI)	P value
PD-1/CXCR3	DLBCL	0.970 (0.910-1.000)	< 0.001b
PD-1/CXCR3	CLL	0.950 (0.880-1.000)	< 0.001b
SIRI	DLBCL	0.820 (0.700-0.940)	0.008a
SIRI	CLL	0.880 (0.780-0.980)	0.002a

^aP < 0.01.

^bP < 0.001.

PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; SIRI: Systemic inflammation response index; AUC: Area under the curve; CI: Confidence interval; DLBCL: Diffuse large B-cell lymphoma; CLL: Chronic lymphocytic leukemia.

Subtype-stratified survival analysis

In subtype-specific analysis, low PD-L1/CXCR3 expression was associated with longer OS in both CLL and DLBCL. High SIRI predicted improved OS in CLL but showed no significant impact in DLBCL (Figure 9, Table 15).

Figure 9 Kaplan-Meier survival curves for overall survival stratified by biomarker levels and subtype in patients with lymphoproliferative neoplasms (n = 90). Panels show programmed death-ligand 1 (PD-L1)/C-X-C motif chemokine receptor 3 (CXCR3) (low vs high) for diffuse large B-cell lymphoma (DLBCL) and chronic lymphocytic leukemia (CLL), and the systemic inflammation response index (SIRI; high vs low) for DLBCL and CLL. Dark blue indicates the group with longer overall survival (low PD-L1/CXCR3 in both subtypes, high SIRI in CLL), and dark red indicates the contrast group (high PD-L1/CXCR3, low SIRI in CLL, and high SIRI in DLBCL, where non-significant). P-values are from log-rank tests, with risk tables below each curve. Mean overall survival values are 285 ± 1.5 months for low PD-L1/CXCR3 in DLBCL, 30.0 ± 1.0 months in CLL, 20.0 ± 2.0 months for high SIRI in DLBCL, and 25.0 ± 1.5 months in CLL. DLBCL: Diffuse large B-cell lymphoma; PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; CCL: Chronic lymphocytic leukemia; SIRI: Systemic inflammation response index.

Table 15 Subtype-stratified Kaplan-Meier analysis for overall survival, mean ± SD.

Biomarker	Subtype	OS (months)	P value
PD-L1/CXCR3 (low)	DLBCL	28.5 ± 1.5	0.015a
PD-L1/CXCR3 (low)	CLL	30.0 ± 1.0	0.010a
SIRI (high)	DLBCL	20.0 ± 2.0	0.120
SIRI (high)	CLL	25.0 ± 1.5	0.008b

^aP < 0.05.

^bP < 0.01.

PD-L1: Programmed death-ligand 1; CXCR3: C-X-C motif chemokine receptor 3; SIRI: Systemic inflammation response index; DLBCL: Diffuse large B-cell lymphoma; CLL: Chronic lymphocytic leukemia; OS: Overall survival.

DISCUSSION

LPNs, including CLL, DLBCL, FL, and MCL, are a heterogeneous group of hematological malignancies with substantial clinical and prognostic variability[1]. In Egypt, the relatively high incidence of NHL, influenced by HCV prevalence and environmental exposures, highlights the need for reliable prognostic biomarkers to guide personalized treatment strategies[2,3]. The present study evaluated the prognostic significance of PD-L1/PD-1 co-expression, CXCR3-associated inflammatory signatures, and systemic inflammatory indices (SII and SIRI) in Egyptian patients with LPNs. Through integration of flow cytometry immunophenotyping with clinical and laboratory assessments, our goal was to clarify the role of these immune-inflammatory markers in disease staging, therapeutic response, and survival outcomes, thereby addressing a critical gap in localized biomarker evidence.

This discussion was organized thematically, integrating hematological, biochemical, immunophenotypic, and survival findings to provide a comprehensive evaluation of the prognostic significance of PD-L1/PD-1, CXCR3, SII, and SIRI. This approach emphasizes synergistic patterns and their clinical and biological relevance.

In our cohort, the mean age was 59.17 years with a slight female predominance (Table 1). CLL was the most common subtype, followed by DLBCL, FL, and MCL (Table 2). More than half of the patients presented with advanced-stage disease (stage III-IV), consistent with patterns reported in Egyptian cancer registries and hospital-based cohorts[11,12]. This distribution likely reflects systemic diagnostic delays common in low-income and middle-income countries (LMICs), where late presentation and limited access to specialized hematology services remain major barriers[11]. Although the sample size was modest, it was sufficient to detect significant differences in prognostic markers and survival outcomes, as confirmed by post hoc power analysis. The inclusion of four distinct LPN subtypes, comprehensive biomarker profiling, and tightly matched controls strengthened analytical power and generalizability. Nonetheless, external validation in larger, multicenter cohorts remains necessary.

Our findings highlight the value of incorporating immune-inflammatory biomarkers into diagnostic workflows, particularly in low-resource settings where molecular profiling is often unavailable. These results are most relevant to populations with a high burden of chronic infections, such as HCV, and restricted access to advanced molecular diagnostics. Validation in larger and more diverse cohorts will be required to establish broader applicability.

The predominance of CLL and advanced-stage DLBCL in our cohort is consistent with previous epidemiological studies in Egypt, where CLL accounts for a major proportion of lymphoid malignancies and is most often diagnosed in older adults[13]. This contrasts with data from Western populations, where earlier detection is more common due to widespread screening, higher health literacy, and timely referrals[14]. For example, the Surveillance, Epidemiology, and End Results registry in the United States reports that approximately 55% of NHLs are diagnosed at stage I-II, largely owing to routine health checks and greater access to imaging[15].

Egypt’s high prevalence of HCV has long been implicated in the pathogenesis of BCLs, particularly CLL and marginal zone lymphoma. Chronic antigenic stimulation from HCV drives clonal B-cell expansion and promotes lymphomagenesis[16,17]. This biological association is reinforced by studies showing regression of HCV-associated lymphomas following antiviral therapy, underscoring its role as a modifiable risk factor[18]. Our findings are consistent with these observations, suggesting that in endemic settings such as Egypt, HCV may serve not only as a risk factor but also as a potential therapeutic target.

The predominance of late-stage presentation in our cohort is consistent with regional Middle Eastern reports. A multicenter study from Saudi Arabia and Jordan found that more than 60% of patients with NHL presented with advanced-stage disease, frequently accompanied by B symptoms and extranodal involvement, closely paralleling our findings[19]. Comparable trends in North African countries further highlight the urgent need for early detection strategies and community-based awareness programs[20].

Collectively, these demographic and staging patterns underscore the need for accessible, cost-effective prognostic tools tailored to local epidemiology. Incorporating immune-inflammatory markers such as SII, SIRI, and flow cytometry immunophenotyping into routine practice may provide a practical approach for early risk stratification in resource-limited settings where PET-CT and molecular profiling are not readily available.

Patients with LPNs in our cohort exhibited classical hematologic and biochemical abnormalities reflecting tumor burden and systemic immune activation. The most consistent findings were anemia, thrombocytopenia, lymphocytosis, hypoalbuminemia, and elevated inflammatory and tumor-related markers, including CRP, ESR, ferritin, LDH, and B2M (Table 3). These alterations are characteristic of active disease and contribute to both staging and prognosis.

Anemia in LPNs arises through several mechanisms, including bone marrow infiltration by malignant cells, impaired erythropoiesis driven by inflammatory cytokines (notably interleukin-6), and increased hepcidin synthesis, which blocks iron mobilization from stores, a hallmark of anemia of chronic disease[21,22]. The elevated ferritin levels observed in our patients (Table 3) likely reflect both iron sequestration and acute-phase reactivity, serving as a surrogate for systemic inflammation[23]. High ferritin has been independently linked to a poor prognosis in lymphoma, particularly in DLBCL and HL, where it correlates with disease burden and cytokine-driven immune dysregulation[24].

LDH, a glycolytic enzyme released during rapid cell turnover, is a well-established prognostic marker in high-grade lymphomas. Our patients exhibited a 5-fold increase in LDH compared to controls (Table 3), consistent with enhanced anaerobic glycolysis (the Warburg effect) and tissue hypoxia in aggressive subtypes such as DLBCL and MCL[25]. Elevated LDH has been included in prognostic models like the international prognostic index and remains a surrogate for tumor mass and proliferative activity[26].

B2M, a component of the major histocompatibility complex class I complex, is another critical biomarker. In our cohort, B2M levels were significantly elevated (Table 3), particularly in CLL cases, where it has long been used to stratify risk and predict time to treatment[27]. B2M reflects tumor cell burden and is influenced by renal clearance; elevated levels have been linked to both poor progression-free survival (PFS) and OS in patients with CLL and NHL[28].

Hypoalbuminemia was prevalent in our patients (Table 3) and can be attributed to systemic inflammation, hepatic protein suppression, and increased capillary permeability. It has been widely recognized as an adverse prognostic factor in lymphomas, often reflecting malnutrition, inflammation, or advanced disease[29]. Elevated CRP and ESR levels further supported the inflammatory milieu in our cohort, contributing to immune suppression and remodeling of the tumor microenvironment.

Comparative studies from Europe and Asia echo these findings. Hamed Anber et al[30] reported elevated CRP and ferritin in aggressive BCLs, reinforcing their role in monitoring disease activity and relapse risk. In a Turkish cohort, Seo et al[31] found that high B2M and LDH levels were independently associated with poor outcomes in DLBCL, confirming their global relevance. Moreover, hypoalbuminemia has been integrated into the Glasgow Prognostic Score, a validated tool in hematological malignancies for survival prediction[32]. Collectively, these hematologic and biochemical alterations delineate disease burden and align with established prognostic systems. Their integration with immune-inflammatory biomarkers, as demonstrated in our study, may provide a more robust framework for individualized risk stratification and treatment planning.

In our study, both the SII and the SIRI were significantly elevated in patients with LPNs, particularly in DLBCL and CLL subtypes (Tables 4 and 6). These indices correlated with disease severity and adverse outcomes, with SIRI demonstrating superior prognostic performance, predicting stage IV disease (AUC = 0.846; Table 8) and independently associating with shorter OS and EFS (Tables 12 and 13). Post-treatment reductions in both indices (Table 5) further support their value as dynamic markers for therapeutic monitoring.

The SII, derived from neutrophil, lymphocyte, and platelet counts, integrates three key components of host-tumor interactions: Neutrophil-driven inflammation, lymphocyte-mediated antitumor immunity, and platelet-facilitated tumor progression. Elevated SII reflects enhanced inflammation with impaired immune surveillance. Neutrophils promote angiogenesis and metastasis through matrix metalloproteinases and vascular endothelial growth factor secretion, while lymphopenia indicates diminished cellular immunity[33]. Platelets contribute by shielding circulating tumor cells and supporting extravasation[34]. The SIRI, incorporating monocyte counts alongside neutrophils and lymphocytes, offers an additional layer of immunologic insight. Monocytes can differentiate into tumor-associated macrophages, which contribute to immune suppression, angiogenesis, and extracellular matrix remodeling within the tumor microenvironment[35]. Thus, high SIRI reflects both systemic inflammation and the recruitment of immunosuppressive myeloid cells, making it a powerful biomarker for tumor-host interactions.

In our cohort, SII was highest in DLBCL and FL, aligning with their aggressive biology, while SIRI peaked in CLL, a disease known for its immunomodulatory and monocyte-driven pathogenesis (Table 6). Interestingly, SIRI showed a paradoxical decline in stage IV cases (Table 7), possibly reflecting terminal immune exhaustion or apoptosis of monocyte subsets during late disease progression. Similar findings have been reported in murine models of lymphoid malignancy, where advanced-stage disease is characterized by impaired monopoiesis and myeloid cell depletion[36].

Global evidence reinforces our findings. In China, a large retrospective study of over 200 patients with DLBCL showed that high SII independently predicted poorer OS and PFS and outperformed NLR and PLR[37]. In Turkey, Gültürk et al[38] reported SIRI as an independent predictor of mortality in newly diagnosed DLBCL, correlating with high international prognostic index scores and extranodal disease. European cohorts similarly demonstrated that elevated SIRI and SII predict adverse outcomes in both indolent and aggressive lymphomas, supporting their applicability across diverse populations[39].

What sets our findings apart is the use of SII and SIRI in a Middle Eastern population, where delayed presentation, chronic infections (e.g., HCV), and limited imaging access often complicate traditional staging. Here, SIRI emerged as a sensitive, non-invasive tool for identifying stage IV disease, suggesting a pragmatic application in resource-limited settings. Its superior AUC (0.846) compared to SII (0.650) and strong correlation with PD-L1/CXCR3 co-expression (r = 0.58, Table 11) further underscore its biologic plausibility and clinical utility. Moreover, the post-treatment reduction in both indices (Table 5) implies their responsiveness to therapy and potential role in early relapse prediction. Such dynamic monitoring is especially valuable in settings where advanced imaging (e.g., PET-CT) is unavailable or unaffordable. In summary, SII and SIRI serve as robust, inexpensive, and clinically actionable markers of inflammation-driven tumor progression in LPNs. Their integration into routine hematological evaluation may enhance early risk stratification, treatment personalization, and long-term surveillance - particularly in under-resourced regions.

Our study demonstrated significantly elevated co-expression of PD-L1/CXCR3 and PD-1/CXCR3 on T lymphocytes in patients with LPNs compared to controls (Table 4). Both markers were closely associated with advanced disease stage (Table 7) and poorer survival outcomes (Tables 9 and 10). Importantly, their levels declined significantly after treatment (Table 5), indicating that these biomarkers not only mirror disease burden but also dynamically respond to therapy. Multivariate Cox regression further confirmed their independent prognostic value for both overall and EFS (Table 12), underscoring their potential utility in immunologic risk stratification.

The PD-1/PD-L1 axis is a key immune checkpoint regulating T-cell exhaustion in chronic infections and cancers. Engagement of PD-1 by PD-L1 induces T-cell anergy, reduces cytokine production (e.g., IFN-γ, interleukin-2), impairs cytotoxic function, and promotes apoptosis of effector T cells[40]. In LPNs, particularly DLBCL and CLL, PD-1 is frequently upregulated on CD8+ tumor-infiltrating lymphocytes, while PD-L1 is expressed by both tumor cells and tumor-associated macrophages, enabling immune escape[41,42].

CXCR3, a chemokine receptor classically linked to Th1-mediated responses, is expressed on activated CD4+ and CD8+ T cells and directs their migration toward IFN-γ-inducible ligands (CXCL9, CXCL10, CXCL11)[43]. In cancer, CXCR3 signaling plays a dual role. While it initially supports effector T-cell recruitment, chronic antigen exposure combined with PD-1 engagement can shift CXCR3+ T cells toward exhaustion, leading to the accumulation of dysfunctional effector memory subsets[44]. This exhausted CXCR3+PD-1+ T-cell phenotype has been described in several cancers, including lymphoma and melanoma, and is associated with impaired effector function and adverse survival outcomes[45].

Our findings reinforce this paradigm. The elevated PD-1/CXCR3 and PD-L1/CXCR3 co-expression observed in patients with stage IV disease (Table 7) likely reflects a dysfunctional immune microenvironment, where chronically stimulated T cells lose effector capacity yet persist due to CXCR3-mediated retention in lymphoid tissues. These exhausted T cells are refractory to reactivation and have been linked to resistance to immunochemotherapy and poor prognosis[46].

Previous studies support the prognostic value of these markers. Kiyasu et al[47] identified PD-L1 expression on tumor cells as an independent predictor of poor prognosis in DLBCL, particularly in non-germinal center B-cell-like subtypes. Xing et al[48] reported that PD-1 expression on peripheral and tissue-infiltrating T cells in CLL correlated with shorter PFS. More recently, Böttcher et al[49] demonstrated that CXCR3+PD-1+ T cells within lymphoma microenvironments show impaired proliferation and cytokine secretion, consistent with functional exhaustion.

Our study is among the first to report PD-1/CXCR3 and PD-L1/CXCR3 co-expression as composite markers in Egyptian patients with LPN. These dual markers provided high discriminatory power for stage IV disease (AUC = 0.872-0.959; Table 8) and retained prognostic value in multivariate models. Notably, their post-treatment reduction (Table 5) implies that they may serve not only as static prognostic markers but also as dynamic indicators of treatment response, with potential relevance for future immunotherapy stratification. Incorporating these immune checkpoint profiles into routine flow cytometry panels could enhance clinical decision-making, particularly in identifying patients who may benefit from PD-1/PD-L1 checkpoint blockade. While such therapies are still emerging in hematologic malignancies, our findings support their exploration in future trials targeting the exhausted CXCR3+ T-cell compartment.

Our study revealed marked dysregulation in CXCR3 expression across monocyte subsets in patients with LPNs, with the most profound changes observed in classical (CD14++CD16-) and intermediate (CD14++CD16+) monocytes (Table 4). Specifically, CXCR3 expression was significantly reduced on classical monocytes, while intermediate and non-classical subsets showed elevated expression compared to controls. This pattern varied by lymphoma subtype (Table 6) and disease stage (Table 7), highlighting a dynamic shift in monocyte polarization and chemokine receptor expression during lymphoma progression.

Monocytes are a heterogeneous population of innate immune cells that play crucial roles in tumor surveillance, inflammation, and tissue remodeling. Based on CD14 and CD16 expression, they are classified into classical monocytes (CD14++CD16-), which are pro-inflammatory and efficient phagocytes; intermediate monocytes (CD14++CD16+), which possess antigen-presenting and cytokine-producing functions; and non-classical monocytes (CD14+CD16++), which patrol the vasculature and participate in endothelial homeostasis[50].

In cancer, particularly hematologic malignancies, monocytes undergo phenotypic reprogramming in response to tumor-derived signals, leading to the expansion of immunosuppressive subpopulations and skewed chemokine receptor profiles[51]. Our finding of reduced CXCR3 on classical monocytes suggests their diminished migratory capacity toward IFN-γ-inducible chemokines (CXCL9-11), which are critical for initiating antitumor responses. Conversely, elevated CXCR3 on intermediate and non-classical monocytes may reflect a compensatory redistribution or chronic activation in the tumor microenvironment[51].

This phenotypic plasticity aligns with prior studies showing that monocyte subsets in patients with lymphoma are functionally altered. For instance, Jones et al[52] demonstrated CXCR3 downregulation in classical monocytes from patients with BCLs, accompanied by impaired chemotaxis and antigen presentation. In murine models of DLBCL, CXCR3 ligands produced in the tumor microenvironment selectively recruited intermediate monocytes, which subsequently differentiated into tumor-associated macrophages with immunosuppressive profiles[53].

Our data also show that CXCR3 expression on classical, intermediate, and non-classical monocytes decreased with advancing stage, particularly in stage IV disease (Table 7). This trend supports the hypothesis that during tumor progression, monocytes become increasingly dysfunctional, lose migratory responsiveness, and adopt a suppressive, pro-tumoral phenotype. Furthermore, CXCR3 expression on intermediate monocytes showed borderline predictive value for stage IV disease in ROC analysis (AUC = 0.720; Table 8), indicating its potential as a staging biomarker. Interestingly, CXCR3 expression on monocyte subsets did not significantly change post-treatment (Table 5), suggesting that monocyte reprogramming may be a relatively stable or treatment-resistant feature of LPNs. This finding is in line with studies by Jakubzick et al[54] and others, who reported persistent epigenetic modifications in monocyte precursors that sustain immunosuppressive traits despite systemic therapy.

Survival analysis revealed that high CXCR3 expression on classical and intermediate monocytes was associated with better OS (Figure 4, Table 9), suggesting a protective role of preserved CXCR3 signaling in monocyte-mediated immune surveillance. These results parallel findings in NHL, where preserved monocyte chemokine receptor profiles, including C-C motif chemokine receptor 2 and CXCR3, were linked to improved T-cell priming and reduced tumor microenvironment immunosuppression[55].

Together, our findings highlight the prognostic and biological relevance of CXCR3 reprogramming in monocyte subsets. The preferential depletion of CXCR3 on classical monocytes and accumulation in intermediate subsets reflects a tumor-induced remodeling of innate immunity, which may be leveraged in future immunomodulatory strategies targeting monocyte-macrophage differentiation and chemokine axis inhibition.

Our study underscores the utility of integrating immune-inflammatory and checkpoint-based biomarkers for enhanced prognostic assessment in LPNs. Among these, the SIRI, PD-1/CXCR3 co-expression, and CXCR3 expression on monocyte subsets independently predicted adverse outcomes (Tables 9, 10, 11, 12, and 13), with SIRI and PD-1/CXCR3 levels emerging as the most robust prognostic indicators in multivariate analysis. This layered approach reflects the complex interplay between systemic inflammation, tumor-induced immune exhaustion, and innate immune reprogramming in LPNs.

The predictive superiority of SIRI over SII (AUC 0.846 vs 0.650; Table 8) may stem from the inclusion of monocytes - a key driver of immunosuppression and tumor support in hematologic malignancies. Monocytes give rise to tumor-associated macrophages, which contribute to angiogenesis, matrix remodeling, and suppression of effector T-cell function. Elevated SIRI, therefore, not only indicates systemic inflammation but also mirrors the expansion of suppressive myeloid populations, which are known to predict poor response to chemotherapy and immune checkpoint inhibitors[56,57].

The PD-1/CXCR3 co-expression on T cells provided strong prognostic discrimination for both OS and EFS (Figure 4, Table 10), particularly in patients with advanced-stage disease. This phenotype, characterized by impaired proliferation, cytokine secretion, and cytotoxicity, represents a hallmark of T-cell exhaustion, as described in chronic infections and cancers[58]. Notably, our study found that PD-1/CXCR3 was a stronger predictor of poor outcomes than traditional biomarkers such as LDH or ferritin. This suggests that immune exhaustion phenotypes may capture disease aggressiveness and host tolerance mechanisms more effectively than metabolic surrogates.

Furthermore, CXCR3 expression on classical and intermediate monocytes independently predicted OS (Table 12). High CXCR3 expression on these subsets was associated with better prognosis, possibly reflecting a preserved capacity for immune cell trafficking and anti-tumor responses. By contrast, CXCR3 downregulation may mark terminal monocyte exhaustion or conversion into non-inflammatory tumor-associated macrophages - a process linked to resistance to chemotherapy and immune evasion[59].

When considered together, these markers represent distinct but complementary immunologic domains: SIRI reflects systemic inflammatory burden and monocyte activation, PD-1/CXCR3 captures adaptive T-cell dysfunction and immune checkpoint exhaustion, and monocyte CXCR3 indicates the integrity of innate immune surveillance. This multimodal strategy aligns with emerging trends in integrative prognostication; wherein immune and inflammatory markers are layered onto classical clinical indices. For example, recent studies in DLBCL have combined NLR, SIRI, and PD-L1 expression into immune-clinical scores that outperform international prognostic index alone in predicting treatment outcomes and survival[60,61].

Our ROC-based analysis (Table 8) and Kaplan-Meier survival curves (Figure 4) confirm the clinical relevance of such biomarker integration. Importantly, many of these markers, particularly SIRI and flow cytometry co-expression indices, are cost-effective, reproducible, and readily obtainable, making them attractive for implementation in resource-limited settings such as Egypt and the broader Middle East and North Africa region. In summary, the combined use of SIRI, PD-1/CXCR3, and monocyte CXCR3 expression provides a robust, immunologically grounded framework for risk stratification in LPNs. This approach may complement or even supersede traditional staging tools in certain clinical contexts and should be explored further in prospective, multicenter studies to validate its generalizability and utility in therapeutic decision-making[61].

The findings of this study carry substantial clinical relevance, particularly in LMICs such as Egypt, where access to advanced diagnostic modalities remains limited. The immune-inflammatory biomarkers evaluated herein, SIRI, SII, PD-1/CXCR3, and monocyte CXCR3, are derived from routine hematological and flow cytometry panels, which are widely available, inexpensive, and technically feasible across most tertiary and even secondary care hospitals. In the absence of PET, next-generation sequencing, or high-throughput gene expression profiling, many hematology units in resource-constrained settings face challenges in risk-stratifying patients beyond basic clinical staging. Traditional tools like the international prognostic index and Ann Arbor staging, while useful, do not capture the dynamic host-tumor immune interactions that profoundly influence disease progression, treatment response, and survival. The incorporation of functional immune markers such as SIRI and PD-1/CXCR3 co-expression may therefore offer a practical bridge between resource availability and precision oncology.

These markers are particularly appealing for LMIC contexts due to several advantages: (1) Cost-effectiveness: SIRI and SII are calculated from complete blood counts, costing under 1 dollar per test in most public health systems; (2) Flow cytometry markers like PD-1 and CXCR3 are part of standard antibody panels in immunophenotyping labs, with no need for specialized infrastructure; (3) Reproducibility and standardization: These indices are based on objective, quantitative values, minimizing inter-observer variability compared to morphologic scoring or clinical staging; and (4) Dynamic utility: Their ability to track pre- and post-treatment changes (Table 5) suggests potential use in early treatment response assessment and relapse prediction. Similar approaches have been validated in other LMICs. In India and Pakistan, several hematology centers have incorporated NLR, PLR, and SIRI into routine lymphoma workflows with demonstrable improvements in prognostication, especially where PET/CT was unavailable[62,63]. A multicenter study in sub-Saharan Africa also demonstrated that inflammatory biomarkers, including CRP and ferritin, predicted survival in patients with lymphoma where molecular diagnostics were inaccessible[64]. Importantly, our study also illustrates the added value of combining multiple low-cost markers, for example, using SIRI for systemic inflammation, PD-1/CXCR3 for adaptive immune exhaustion, and monocyte CXCR3 for innate surveillance integrity. This layered approach, while inexpensive, provides a more nuanced risk portrait than any single biomarker, particularly in CLL and DLBCL, which dominated our cohort.

From a policy and public health standpoint, the adoption of such biomarkers may also assist in resource triage, for instance, identifying high-risk patients for closer monitoring, hospitalization, or referral to specialized centers. This is particularly critical in overburdened oncology systems, where bed availability and chemotherapy access are constrained. In conclusion, the immunologic and inflammatory biomarkers identified in this study represent scalable, accessible, and biologically meaningful tools for the prognostic assessment of LPNs in Egypt and similar LMIC settings. Future research should focus on external validation across diverse populations, integration into prognostic models, and implementation science frameworks to optimize clinical adoption[65].

This study is among the first to comprehensively evaluate PD-L1/PD-1 co-expression with CXCR3 and systemic inflammatory indices (SII and SIRI) in Egyptian patients with LPNs. The high incidence of NHL in Egypt, partly attributed to endemic HCV infection and environmental exposures, underscores the need for localized biomarker research tailored to the region’s unique epidemiological landscape. Unlike prior Egyptian studies, which have largely focused on descriptive statistics or clinical outcomes, our work integrates advanced immunophenotyping with systemic inflammation-based indices to propose a novel, biologically grounded prognostic panel. The identification of PD-1/CXCR3 co-expression as a powerful predictor of stage IV disease in both DLBCL and CLL, along with the dynamic stage-specific behavior of SIRI, contributes region-specific data that may refine risk stratification in this high-burden population. By addressing both immune exhaustion and inflammatory dysregulation, our findings provide a contextualized framework for precision oncology in resource-limited environments. These results suggest that co-expression of PD-L1/PD-1 with CXCR3 and elevated SIRI may serve as clinically actionable biomarkers to guide prognosis and potentially inform therapeutic stratification in LPNs.

This study had several limitations that should be taken into account when interpreting the findings. First, it was conducted at a single tertiary center in Egypt, which may limit the generalizability of the results to broader populations or different healthcare systems. Although a range of lymphoma subtypes and disease stages were included, larger multicenter studies are needed to validate the prognostic performance of SIRI, SII, PD-1/CXCR3, and monocyte CXCR3 expression in more diverse cohorts. Second, while our flow cytometry analysis provided valuable information on immune checkpoint and chemokine receptor expression, we did not perform functional immune assays such as cytokine profiling or in vitro exhaustion analyses, which could have enriched our understanding of the biological mechanisms underlying immune dysfunction in LPNs. Third, our follow-up period was relatively short, which may limit our ability to draw definitive conclusions about the long-term prognostic value of these biomarkers, particularly in predicting relapse or progression beyond first-line treatment. Additionally, molecular profiling was not available in this cohort, precluding integration of immunologic markers with key genetic alterations such as v-myc avian myelocytomatosis viral oncogene, B-cell lymphoma 2, or tumor protein p53 status. Therapeutic heterogeneity, while reflective of real-world clinical practice, may have also introduced confounding effects, as patients received different treatment regimens based on subtype and available resources, potentially influencing biomarker expression and outcomes. Furthermore, while we propose that these biomarkers are affordable and suitable for resource-limited settings, we did not formally assess their implementation feasibility in low-resource healthcare environments. Future studies should aim to validate these markers prospectively in larger, multicenter cohorts, ideally integrating immunologic, genomic, and radiologic data into unified prognostic models. Additionally, exploring the use of these biomarkers to guide treatment selection, particularly in identifying patients who may benefit from immunotherapy, and evaluating their cost-effectiveness and applicability in routine practice across diverse health systems, will be essential for translating these findings into clinical impact.

These findings not only validate the prognostic role of PD-L1/PD-1-CXCR3 co-expression and systemic inflammation indices in Egyptian patients but also provide a model for integrating immune phenotyping into routine hematology-oncology practice. This integrative biomarker approach could be readily adapted in other resource-limited settings, offering both prognostic precision and cost-effectiveness.

CONCLUSION

Our study provides robust evidence that combining immune checkpoint co-expression markers, such as PD-1/CXCR3 and PD-L1/CXCR3, with systemic inflammatory indices like SIRI offers a powerful, accessible approach to prognostic stratification in LPNs. These markers not only correlated with disease stage and treatment response but also independently predicted survival, highlighting their potential utility in both clinical risk assessment and therapeutic decision-making. Importantly, their availability through standard laboratory and flow cytometry platforms makes them feasible tools for implementation in low-resource settings such as Egypt, where access to advanced imaging and molecular profiling is often limited. Together, these findings support the integration of immune-inflammatory biomarkers into routine hematologic practice to improve patient outcomes and enable more personalized care across diverse healthcare environments.